• Intro to functional genomics assays to study transcriptional regulation and epigenetics
  • ChIP-seq and beyond
• R Bioconductor infrastructure for analysis of genomic data

• Practical steps for analysis of ChIP-seq
  • Obtain the data
  • Align reads to the genome
  • Call peaks
  • Identify reproducible peaks between replicates
  • Find motifs

• Public repositories
  • Gene Expression Omnibus (GEO)
  • Sequence Read Archive (SRA)
• Consortia
  • ENCODE
  • Roadmap Epigenomics
  • etc.
• Your own data
  • Integrated Genomics Operation (IGO)

• https://www.ncbi.nlm.nih.gov/geo/

• https://www.ncbi.nlm.nih.gov/sra

• https://www.encodeproject.org/

• https://github.com/qbio-gsk/qbio-gsk/tree/master/2019/lab8

• File names for most large files, as placeholders

• https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE55773

• SRA toolkit

• Download one file:

fastq-dump -I --gzip -A SRR1186971

• Download all files:

#!/bin/bash

for fileid in SRR1186971 SRR1186973 SRR1186974 SRR1186976
do
    fastq-dump -I --gzip -A $fileid
done

https://github.com/qbio-gsk/qbio-gsk/tree/master/2019/lab8/data/chipseq-fastq

• What is sequence alignment
  • Classic problem in bioinformatics
  • Align (DNA, protein) sequences for evolutionary analysis
  • BLAST
• Tools to align millions of short reads to large genomes
  • Bowtie2, BWA, STAR

• Align reads from one sample:

bowtie2 -p 10 -t --no-unal -X 500 -x mm10-bowtie2index/mm10 -U SRR1186971.fastq.gz >SRR1186971.sam 2>.bowtie2.SRR1186971

• Bowtie2 output:

Time loading reference: 00:00:00
Time loading forward index: 00:00:01
Time loading mirror index: 00:00:00
Multiseed full-index search: 00:04:06
24505004 reads; of these:
  24505004 (100.00%) were unpaired; of these:
    2832435 (11.56%) aligned 0 times
    13902816 (56.73%) aligned exactly 1 time
    7769753 (31.71%) aligned >1 times
88.44% overall alignment rate
Time searching: 00:04:08
Overall time: 00:04:08

• Often in bioinformatics: data stored in tab-delimited text files, command line tools to manipulate them
  • samtools, bedtools, deepTools, etc.
• Sequence Alignment/Map (SAM)
  • SAMtools to manipulate SAM files

• Tools for summary statistics and quality metrics
  • FastQC
  • samtools stats
  • MultiQC
• FastQC

• Align all samples (bash script):

https://github.com/qbio-gsk/qbio-gsk/blob/master/2019/lab8/align/run-alignment.sh

• Rename samples for convinience:

https://github.com/qbio-gsk/qbio-gsk/blob/master/2019/lab8/align/rename-samples.sh

• You will get files shown in this folder:

https://github.com/qbio-gsk/qbio-gsk/tree/master/2019/lab8/data/chipseq-bam

• Considerations
  • control: input DNA, non-specific antibody, genetic controls
  • paired-end or single-end sequencing
• Tools
  • MACS2
  • SPP
• Reproducible peaks (ENCODE standard)
  • Irreproducible Discovery Rate (IDR)

https://github.com/taoliu/MACS

macs2 callpeak -t foxp3_rep1.bam -c foxp3_input.bam -f BAM -n foxp3_rep1 -B --SPMR --outdir peaks-macs2/foxp3_rep1/ -g mm -p 0.1 --keep-dup 'auto' --call-summits 2>.macs

• Peaks output in narrowPeak format (extended BED)

idr --verbose --samples foxp3_rep1_peaks.narrowPeak foxp3_rep1_peaks.narrowPeak -o foxp3_rep1_foxp3_rep2.narrowPeak --log-output-file $idrdir/$prefix.log.txt --plot

• Results of peak calling and selecting for reproducible peaks:

https://github.com/qbio-gsk/qbio-gsk/tree/master/2019/lab8/peakcalling

• (Some) Proteins bind to DNA in a sequence-specific manner

• Motif to represent and model sequence specificity
  • Positional Weight Matrix (PWM)
  • Motif logo

• Integrative Genomics Viewer (IGV)

• UCSC Genome Browser

• File formats
  • bedGraph
  • wig
  • bigWig
• Tools to generate bigWig
  • bedtools genomecov, followed by UCSC tools to convert bedgraph to bigWig
  • deepTools